Wednesday, July 02, 2008

Is the OntologyWorks Knowledge Server a Database?

I was looking for companies that are working with technologies related to Semantic Web. I noticed that a few members in the RIF (Rule Interchange Format) workgroup at w3c are from OntologyWorks.

define themselves as,
".. a product company offering a broad suite of semantic technologies including deductive information repositories (our Ontology Works Knowledge Servers), semantic information fusion and cost effective semantic federation of legacy databases, ontology-based domain modeling, and management of the distributed enterprise."

It wasn't unusual for them to justify their existence, considering the fact that there are a number of disbelievers as far as the ability of Semantic Web concerned. One question that they answered was the thing that whether their capabilities were any different from companies offering typical database management system based services. Following is their
"Our information repositories are called Ontology Works Knowledge Servers to distinguish them from traditional database management systems or “databases” as these are usually called. Our Knowledge Server Family of ontology-based deductive systems do all that a database does but also do so much more that. They are a change of state in database technology, as from water to steam. We also apply our semantic technology capabilities to other information management systems, e.g. semantic legacy database federation, semantic enterprise mission management, semantic enhancement of existing databases and other applications.

In essence, traditional databases contain models of the world that are truncated and mute on many issues. They do not do inference over the model and cannot handle 3-place and higher relationships as our systems can. The emphasis in traditional databases is on storing lots of data and getting some of it out fast when needed. They were built so that users could ask for something, not about something. Traditional systems are brittle, expensive to maintain, and can't provide the answers the modern enterprise needs.

At our end of the semantic technologies spectrum we can describe the knowledge domain you care about with full descriptive power - with temporal understanding, n-place relationships and a powerfully descriptive controlling logic that allows complex inference over the model.

The genius of Ontology Works is that we can instantiate this complex, expressive model (ontology + controlling logic) of a knowledge domain in a deductive system - the eXtensible Knowledge Server (XKS) that gives excellent query performance for complex query whose answers have very high value to the enterprise. Our Knowledge Servers give the user real knowledge discovery, not just data retrieval.

The question then is not, “Is it better to have these highly expressive information systems or go back to mute data models?” The question is, “Yes, this expressiveness is an undoubted good, but is it efficiently computable.” With Ontology Works the answer is “Yes, it is efficiently computable.” Consequently, the economics of implementing our systems is very favorable."

No comments: